ancova post-hoc comparisions of slope

[whewa]

I was wondering if there was a way that I could do tukey's pairwise comparisons after testing for parallel slopes using sas.

my sas code to test for parallel slopes is:

proc glm;
class set;
model actual = set estimate set*estimate;
run;

[Mean Joe]

Is this what you want?

proc glm;
class set;
model actual = set estimate set*estimate;
lsmeans set estimate / tukey;
run;

[whewa]

I would like to test the slopes and not the mean. would that test the slopes?